The reaction of epichlorohydrin with trimethylammonium hydrochloride in an aqueous medium to form 3-chloro-2-hydroxypropyl trimethylammonium chloride is well known. The resulting processed aqueous solution is commonly used to manufacture cationically active starch products. However, such an aqueous solution contains substantial quantities of by-products which are non-reactive in the starch manufacturing process, including, for example, 1,3-bis(trimethylammonium chloride)-2-hydroxypropane, and 1,2-hydroxypropane-3-trimethylammonium chloride.
These non-reactive impurities are not easily removed and can pose serious disposal problems to the cationically active starch manufacturers. These impurities can also cause a higher level of FDA extractable quaternarys in paper manufactured with non-washed cationic starch.
Another advantage to removing these impurities is the desire to transport more active 3-chloro-2-hydroxypropyl trialkylammonium chloride per pound of aqueous solution. Since the amount of total solids in the aqueous solution controls its freezing and crystallization point, the removal of these impurities allows a higher concentration of 3-chloro-2-hydroxypropyl trialkylammonium chloride per pound while effectively avoiding the freezing and crystallization point.
As far as Applicant is aware, there has been no method heretofore devised for removing these non-reactive impurities. The only method of which Applicant is aware which can be employed to obtain 3-chloro-2-hydroxypropyl trimethylammonium chloride free of substantial non-reactive impurities is described in U.S. Pat. No. 4,450,295. According to the method described in that patent, an aqueous trialkylammonium chloride solution was first prepared. To this solution was then added a solvent which formed an azeotrope with water, the azeotrope having a boiling point below that of the solvent alone. In addition, the solvent had to be immiscible with water and non-reactive with epichlorohydrin. The azeotrope of the solvent and the water was then distilled off leaving a dispersion of the trialkylammonium chloride in the organic solvent. Thereafter, the trialkylammonium chloride was reacted in the solvent with epichlorohydrin. Such a method was not generally desirable because of the great amount of energy required to distill the water-solvent azeotrope. In addition, existing facilities for manufacturing 3-chloro-2-hydroxypropyl trimethylammonium chloride were not easily retrofitted to employ the process.